Infants of mothers who remain depressed beyond their infants' first few months develop depressed behavior during early interactions and experience later developmental delays. The purposes of this research are: 1) to find predictors of the mother's continuing depression during the first few months; 2) based on those predictors identify those mothers and infants that will need early intervention for those dyads. For the study on identifying risk factors for continuing depression (first 2 years) a number of infant measures (sleep/wake and face-to-face interaction behavior, temperament, heart rate and catecholamines/cortisol) and maternal measures (self-reported depression, anxiety, stress and social support, interaction behavior, EEG, EMG and catecholamines/cortisol) will be monitored at birth, 3 and 6 months in 200 dyads. Using a regression model a cumulative risk index will be formed on the birth and 3 month measures that predicted to continuing depression at 6 months. In a second sample of 200 depressed mother-infant dyads the 80% expected to remain depressed will be identified based on the high risk index score and randomly assigned to a prevention/control group at 3 months in an attempt to prevent infant depression by reducing depressed maternal behavior during the "critical" 3- 6 month social interaction period. The interventions are designed to reduce those maternal behaviors that are likely to contribute to infant depression. They include reducing depressed mood in the mothers (by music mood induction/relaxation therapy), enhancing sensitivity to infant cues (by infant massage and imitation/interaction coaching) and reducing attributions of infant behavior (by video and auditory feedback). The effects of this prevention on infant mood, behavior and development will be assessed across the first two years.